Roel Hacking

PhD Researcher · Computational Illumination Optics · TU/e Eindhoven

I am a PhD researcher in the Computational Illumination Optics group within the Department of Mathematics and Computer Science at Eindhoven University of Technology (TU/e). My work sits at the intersection of machine learning, computational mathematics, and optical design.

My PhD research focuses on developing neural network–based methods for solving the Monge–Ampère equation and designing freeform reflectors for illumination optics. I combine physics-informed neural networks with classical numerical methods to push the boundaries of what's computationally achievable in optical design for both point and finite-size light sources.

Before my PhD, I completed a double Master's degree at Radboud University — in Artificial Intelligence and in Cognitive Neuroscience — and a Bachelor's in Data Science at Maastricht University. My earlier research ranged from prosthetic vision and COVID-19 detection to neuromorphic computing and brain-computer interfaces.

Research interests physics-informed neural networks · partial differential equations · freeform optical design · scientific machine learning · inverse problems · computational imaging

Publications

  • A neural network approach for solving the Monge–Ampère equation with transport boundary condition R. Hacking, L. Kusch, K. Mitra, M. Anthonissen, W. IJzerman 2025Journal of Computational Mathematics and Data Science, 15, 100119
  • Hybrid neural and deconvolution approach for finite-source reflector design R. Hacking, L. Kusch, K. Mitra, M. Anthonissen, W. IJzerman 2025EPJ Web of Conferences, 335, 02011
  • Neural-network based reflector design for finite 2D sources in far-field illumination R. Hacking, L. Kusch, K. Mitra, M. Anthonissen, W. IJzerman 2025Freeform Optics, FrW4A.4
  • Pyrimidine metabolism and related diseases R. Hacking, D. Slenter, E. Willighagen, M. Summer-Kutmon, I. Hemel, et al. 2025

Experience

2022 – present PhD Researcher Eindhoven University of Technology · Computational Illumination Optics

Developing machine learning methods for illumination optics design, solving the Monge–Ampère equation with neural networks, and designing freeform reflectors for finite light sources.

2021 – 2022 Graduate Researcher Neural Coding Lab · Radboud University

Developed convolutional neural networks to evaluate prosthetic vision quality, including a differentiable Canny edge detector for automatic parameter optimisation.

2020 – 2021 Graduate Researcher Diagnostic Image Analysis Group · Radboud UMC

Built ML methods combining chest CT scans and clinical features for COVID-19 detection, using gradient boosting and CNNs with techniques to handle missing data.

2017 – 2019 Data Scientist CBCLab · Maastricht University

Developed data processing pipelines and ML approaches for 3D microscopy image fusion using dual-inverted Selective Plane Illumination Microscopy.

2015 – 2017 Backend Developer PNA-group · Heerlen

Developed the knowledge management software Cognitatie, contributing to features that led to adoption by the Dutch Rijkswaterstaat.

Education

M.Sc. Artificial Intelligence: Cognitive Computing Radboud University · 2022 · cum laude
M.Sc. Cognitive Neuroscience: Neural Computation & Neurotechnology Radboud University · 2022 · cum laude
B.Sc. Data Science and Knowledge Engineering Maastricht University · 2019 · summa cum laude
Gymnasium Graaf Huyn College · 2016 · cum laude

Contact

I'm open to conversations about research collaborations, industry opportunities, and positions in machine learning, scientific computing, or computational engineering.